MR May 19hl

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

ItrIIRTIMI{ REP()RTORIGINALLY. ISSUED

May 1941 as

Memorandum Report

A FLIGHT INVESTIGATION OF TEE BO_DARY-LAYER

CHARACTERISTICS AND PROFILE DRAG OF THE

NACA 39-215 LAMINAR-FLOW AIRFOIL AT

HIGH REYNOLDS NUMBERS

By J. W. Wetmore, J. A. Zalovclk, and Robert C. Platt

Langley Memorial Aeronautical LaboratoryLangley Field, Va.

WAS_IGTON

NATIONAL AD_[SORY COMMITTEE FOR gEI_ONAU_ICS

_MOE&J_r_,_ REPOET ....

for the

A]_ny Air Corps

A FLIC}_f Ii_VESTIC-ATIONOF THE BOT_NDARY--LAYER

r" " c_.rq C_CHARgC_EhlolICo AND PROFILE D]_AG OF TI_

' NACA 35-2--]5IAMIN._-FLOW A.Ii_FOI!.AT

HIGH I_EYNOLDS _US_f_ERS

B 2 J. W Wetmore J. _. Zalovcik, and Ro'berI_C. Pl:_t_

S[_MAP_Y

Tests have been conducted in flight to determine theboundary-J.ayer charaator_.stics and the profile drag of theNACA 35.-2!5 airfoil section at high.Reynolds numbers. Thesetests were made on a test panel of 17-foo*s chord mounted on

the ].eft wing of a Douglas B-18 airplane Just outside of thepropeller slipstream. Tests wore made to determine the tran_-sitioll points and the bo]mdary-.layer velocity profiles forvarious surf&co and power conditions over a range of airplanelift coefficients from 0.20 to 0.!!.6for which the range of

corresponding Reynolds numbers was 30,000,000 to ?O,000,000.The profiie_-drag coefficient of the panel was determined forthe best surface condition both with power on and with the

eng_.nes and propellers _topped o¢er a range of airplane liftcoefficients from 0.2J to 0.32 with a Reynolds number range

of 32,000,000 to 16,000,000. In addition, the profile dragof the u]_por surface alone was debe._mined for the s__,e powerand s_cface condition and over approximately the same rangeof airplane lift coefficients and Reynolds nmnbers.

With the best surface condition and the left engine

stopped• the laminar boundary la?Ter was maintained to 42.4 ?or-cent of the chord on the _py_er surface at a lift coefficient of0.220 and a Reynolds number of 26,700_000. The results of thetransition tests indicated a reduction of about 3 percent ofthe chord in the isminar-.-flowrun over the upper, surface due

to operation of the engines and propellers. As a result of'reducing the indicated amplitude Of the transverse waves onthe upper s_n'fakcefrom 0.005 to 0.00! inch, the transit_on

., point moved back from. about 3?.5 to about 42.5 p0rcent of thechord.

2

The velocity surveys in the laminar boundary layer indicated

that values of boundsr_y--layerReyno.lds nt_ber R 5 (base8 on thed_s_ance above the surface at which the dyn_lic pressure in the

boundary ls.#er is one-half that just outside the bo_nda.ry layer)e_:ceeding CO00 are attainable in f?.ight on suitably designed an@.csrefully finished airfoils.

The profile-drag coefficient of the test panel w_th enginesstopDed was feu.ud to rema.in substantially constant at a valueof about 0.0045 for fl._ght conditions ranging from an airplanelift coefficient of 0.21 and. a corresponding Reynolds nlm_,_erof

about 30,000_000 to a li_t coeff.icient of 0,32 and a Reynoldsm_mber of 24,000_000, Over the same rsnge of conditions theprofile-_.rag coefficient of the upper surface alone varied fromabout. 0.0022 at the lowest lift coeffAclent i,ested to 0.0025 at

the hi_hest lift coefficient. With both engines operating atftull throttle the d:Jragcoefficient due to both surfaces and thatdue to the u_er surface alone were both increased on the orderoi"8 to i0 percent,

The results of the bests indicate the desirability for

co_.tlnued fli,zht resea:_._chon airfoils at l_arge scale to supple-merit the development work of the tu_nels.

INTRO_CCTION

During the earlier sta_oos of the Committee's work on thedevelopment of l,.-_inar-flo_,,_airfoils (reference i)_ it wasfound that by su.ita:o!ydes_.gning the profile of an airfoil a_ - •b]_.a._ora...eor accele-_:atingpressure 2_ra__ientcould, be maintainedover as much as 80 Dercent of t]:',echord back of tlhe leading edge.Tests of some of these airfoils in the wind. tunnels and in fligh.t

showed that, within the lower flight range of Reynolds n_Ibersthe laminar bound.a,rylayer extended as far back as 80 percentof the chord, from the leading edge., with the result that theprofile drag was extremely low.

In the higher Reynolds number ran_es, say, above 20,000,000,it was expected, t_at other methods mig_'Itbe required to obtainthe desired, extensive laminar bo_ndary layers and resultingextremely low drags. The present investigation was undertakenwith the object of investigatin_ methods of prolongJ.ng thelaminar flow at high Reynolds n_bers and to give data forcomparison with wind-tunnel data. Consequently, a suitable wingwas chosen with these objects in view rather than with thisobject of choosing an optimum section for any particularpractical application.

3

This report represents results of the tests of the 91ainairfoil. These tes bs covered a range of Reynolds numbers

between 20,000,000 and. 30,000,000 and included variations _npowe:_ condition and surfece condition. An investigation ofthe effect of section slots for boundary-layer control will

be covered in a subsequent report.

The tests were made with a B-18 airplane which was madeavailable for this project by the A]n_y Air Corps.

APPAP._TUS

The Douglas B-!8 airplane is a bimotored_ fully cantilever,midwing monoplane with a wir:g area of 958.6 square feet and adesign, gross weight of 23,200 l_ounds. It is powered with WrightCyclone R-.1820--45engines (810 horsepower at oi00 rpm and8700 feet) fitted with 3--b!ade propellers having a diameter ofii feet 6 inches. Hamilton Standard, hydraulically controlled,constant-speed propellers are normally used on this airplane,but, for most of the present tests_ they were replaced by Curtisselectrically controlled full-feathering propellers in order thatthe engines could be stopped during flight. The weight of theairplane as flown was approximately 22,000 pounds.

A test panel having the NACA 35-215 airfoil section (table !)was mounted on _:.heleft wing of .the airplane. The chord of thepanel was 17 feet and the span was I0 feet at the leading edge,tapering to 5 feet at the trailing edge. It was constructed oflaminated white pine in tho form of a hollow shell w._.thwallsabout 2 inches thick; the outside profile was accurately shapedto templet size. The surfaces were sprayed with several coats

of lacquer base filler and rubbed down with various grades ofwater cloth, the final finish bein_ obtained with a No. 400water cloth, The panel was supported ou the w_ng by rubber padsrunning along the top and bottom of the wing spars and was securedin place by mear_s of steel straps. The position of the panel wassuch that the inboard end of the leading edge was about i footoutboard of the propeller disk, the leading and trailing edgeswere normal to the plane of sy_mmetry of the airplane j and.thepia,ne of chord lines coincided approximately with the plane of

4_ ,_chord lines of the wing. The panel was _.<_iredinto the w_.ng_bymeans of fabric stretch,sd taut over a wooden framework. The

weight of the pane], and fairin<_ was 1394 pounds; satisfactorylateral balance for all conditions of flight was obtained b,yremoving all. fuel from the left-wing tanks and adding 350 poundsof ballast in the right wing tip. Figure i is a photograph, ofthe test pane], mounted on the wing; its dimensions and locationare shown in figure 2.

The upper surface of the panel was refinished severaltlmes durir.g the course of the tests so that various su:rface

I

con&ibions are represented in the res'_Lts, An index of thesurfa,co wavln,_ss, i, e.., t_,9 magnitude of the bran_verso waves__,zasobtained by measurin_ the cu_Tature variation along thesur_:e_e by me_:_s ..the de_ice _ho_._ in flg_re 3, Fin_.shin_._th_ lower surface w_-,,sre'dud to be v,e.rj"_-d_fflcult...... so that no

a.tte_q_twas made to ",-'efin.'.sh_.tan& no waviness meas-_ments_,Teremade on it ':w,_econdition of the lower st_.:fa(,ethrough ....

out the invesT.,ig_:,_tionis believed to have been 8,bout the sameas the Initla,1 _.-o",_'z.,.,..,_,ion of th,3 up-.or_r.... surface,

Free-.stresl_sua,,tlc and to,,_.l pressures were measured bymeans of static-- t_nd tota].-p:2essurotubes which were calibratedwith a static bead suspended bellow the alrplano.

The characteristics of the bot_dary layer were determinedby me,e/_se_ther of 5-hubs or 2,-tube racks. The 5-,-tuberackswere each composed of a s-[;at!c-pressuretube and fo_r total--pressure tubes azrange& to :meas,e_ethe static pressure Justoutside the boundary layer und the tobal pro_s_re clove tothe surface _d at various dist_nces above the surface within

tn_,2 wore used to determin_ the velocity-theboundary layer; _ "profile of the boundar3_ la_-er, In. cases where it was desiredto detem_mine only the poi;it a%t_,.zhiohtransition occurl_d, the2 +_"'_=racks, each consi_t_.ng of a stM;Ic tube located Justoutside the boundary layer and a totai-presm_e tube .l..ocatedclose to the s_mfaoe, were used.

Wake-t?ressure suz-._eysfor the determinatlon of profi].edrag we','eaccomplished, by me_uns of a bank of o_--,:.ptotal-pre ss_reand 6 static-press_-'e ouoos located 12 percent of the ch_'dback of the trsiiing edge on the ps,nel center llne sm,Lextending•through the_entire wake. The totai--pr_sst_.etubes were spaced0.60 inch ap___rt. A ba_.,_of tubes consistin_ of 21 total_ressuretubes, spaced 0.25 inch apart, and 3 sbatic-pressure tubes,mounted at the center of the tr_:_ilingedge and extending only

through the upper " "°" _sur._ac,,,wake _._s used for the determination

of the profile drag of the upper s_rface alone.

All pressures were meas_red, by means of a multipie-t_bealcohol manometer and were recorded photographically.

TEc_To

Boundary--layer m_as&_ments were r_.de On the upper surfaceof the test panel c_or s.re_K_e of airplane lift coeffici_ntsfrom about 0.20 to 0.4.6; bh8 ra_4;o of corresponding Reynoldsnumbers was ..zcm e_buut 50_000,C<kO bo 20,000,000. S_veralconditions of the p_mei s_,,u'face,as indicated in fig_:_e 4, _idvarious po_er con.ditlons were in'vest_gate_l. T]:_,_po_,._orconditionscovered ware as follows: both engines full throttle; both

e1_gines idling; !eft engine stopped, :right engine full throttle;l_ight engine st,opp_d_ left engine full throttle; both engines

stopped. Only a fow tes%s _ere made on tho lower surface ofthe panel because of its inferior condition.

The profile drag due to both r_urfaces e_td that due to theupper surface alone _as determined with 'the panel surfaces in.the final _ondition and for two po_er cond.itions: both enginesat :Full throttle and both engin_,_sstopped° 'l'h'eprofile-_.ragmeasuremonts cover.sd a ran@_o of airplane lift coefficielzts from0.21 to 0.32 with a renge of corresponding Reynolds n_borsfrom 32,000,000 to 24,000,000.

Inasmuch ss it was necessary to dive the airplane in orderto attain th,3 low lift coefficients desired., the relative lag

of the various pressure tubes and lines was determined byspecial tests and the res_f£ts were corrected eccor&ingiy.

RESUI_TS '

Results of the investigation ai_ presented .infigD1res 5 to i0and in tables II to V. In figure 5 the distributions of press&re

coefficient, S, .(S=q/qo), ovor bhe forward parts of the surfacesare shown. All experimental points in li.gure 5 are for positionsalong the center line of the upper end lower s,tu"_'acesof the tost.psnel and were determined by _ue,sn&s of the bouoAary-layer rackS.

. Transition results are presented _n tables II.and Ill for fo_st_rface conditions as shown in figure )_,and for various engine

az_.propeller conditions. The ranges of lift coefficient a_dRe_naolds number covered in each test rmu are included in additionto the partictt!ar lift coefficients and Reynolds numbers at whichtransition occurred. The method of determining the conditionsfor tra_tsition is indicated in figure 6. In figures 7 a_d 8 thevelocity distributions in the L_tinaro-bo_dary layer are shownfor various chor&wlse and lateral positions on the upper and

6

lower stuffaces _s plots of u/U against -Y-,FB-,whers u isC

tlte " _q ....,-el,city _i_nzn th_ botu_.da,rylayer, U is the velocityJust oul.,s=cLe_ _ the bo'uz.a_,ry...... Zayer_ y is the d.is0ance _'rom thesurface at which u is measure&, c; is the panel chord_ andR in the Reynolds n_._r_fberi:at ...._, of the p_._ol chord and thefree-sbre_ velocit_r_ this me%hod of •olottlng e!iminatos the

...... ' -_- theeflec, b of ve_ria_ons _.n R_<ynolds number. Values of _5"

boundary-layer Re2nolds nambor in terms of U _nd of the valu_of y at _.,b..c,hu/J 0.707, are lists.3,in table IV forvarious condit_.ons under which t_.an,_b_on to tu_rbu!.entf].ow

._ ...... p_of.-].e--d.._gcoefflcients for bothwas mr'obabl.$ _r_mi.nent. ,m_...... _,-surfaces and for the Vgper s_'face a]one arm given in fig_ros 9

and !0, respectively, and in table V.

DISCO_$IO}f

The pressure distribution over the fo_ard 53 percent of thechord on the upper _ "sur±s,c_ an& over 40 percont of bho oh.oralonthe !o_{er s_:,rfacewa_ det_.t.m,ined .fz....mthe static-;oressure meas_'e--m<.L_ obtaiD.ed._.¢iththe botu%dary--iayor racks. Inasmuch as the

Section lift coefficients c_ mould not be _ ,a..,,uabed _i_houtpressuro-distributio:a &ata ever th,9 entire panel chord_ theres_.Lts of the inves[,,Lgationare presented in rele,,tionto the

airplane lift coefficient CL. A span_,_isevo,_iab",.onin th_

surface press_n_es indicated, ths.t the section lift coefficientvaried on _ne or&or of 4 or 5 1_,ercentover the _......_,_.,,<_e,of span_'isepositions covePed in the ' "_.. .. ,,es_s_ b_ing hi_host inboard and lowest

• _Oe£ l ioien'ooutboeor& of the pa.n_], cen.ter line The section lift _ -.........at the center of the test pauel is ostimated to be Eoout 0.90 ofthe airp!aue lift coef..lom_nt

Th:3 experimental pressure distr!buti0n She%u% in figure 5was obta,inod at an airplane lift.•coefficient of 0.238 so tha±,%he section lift coefficient _¢as probably abouJc 0..22 as comparedto the value of 0.20 at _L..mchthe ,:.izfoilis designed to operate.This sm_dl ......._ _ _" ..ct]._]._r_n.uein lift coefficient "_¢OUld probably notmaterially a:£::ect the sh_,_pesof the cu_ves, E:'hem.inimwa pressureon the upper surface In Shown to occttr at about 4-5percent ofthe chord.

The transition conditions s_nnma.rm_edin tables II and I!l_' -_ chord,_.'is,_are define&, as the conditions at which, for a _zv_n

posit.t,-;n,a olmgb..,_e._?.,_.;-_h.._efrom the given .-o.I_ coefficlont .....Roynol&s ntuub_r combination wo[J_d,cause trar.sitlon from ie_inar

7

to turbulent flow. '2he transition wo,s ,generally well defined byan abrup_ rise in th_ velocity close to tho s_m'face as i!!ustrat_din figure 6.

t_..,e transition r_t_l._s for the various con-Comparison of i_t. o , .. - -

di_.ons _es_ed _s :.featherm_ce-_taln in some cases owing to thefact tDat th,-_:r_is no f_._.edre..aolon b_tween airDiaue Iift

coefficient _.nd_' _ _ _ _. , ...._e_,no._.donumlg_r; 4 e. for a Qnazii_itativoe-¢aluation of the effect, for ex_ma?lo_ of the po;¢er or stlrfacecondition on the ext_nt of the ],mnin_._,r--botu_clarylayer; corn--

I"" ( "pa_is)n sho_J-].dbe ma'].eat the sam_ lift coefficient and atthe s_me Rej_.o].dsnumber. There a._"e,however, several con--cluslons !.nd.Ioatedby the res_].tso With the best s_,,u'facecondition tested (condition D, iio. -,-)and with the. ±e._._ enginestopped the _em.-_.narboundary layer we_smaintained to' _.2.4osrcentof the chord on the upper surface_ As sbow_, in table !!, tram.-smolon was observed a_;;this station au several different combi-nations of CT and R owing to the __avoidsfole "_ariation in :

the rel..a_;_.on of R to CI, 'bet_een cliff errant "best ru_.,s. At.... .w_20wl_ich most ne_: y ap,oroachesan airplane lift coefficient of 0 " --i

the desi_ lift coefficient of th_ p_-anel (c_ = 0.20), %heRe_ol..ds nu_,_berfor tre,usition _,t 42.4 percent of the, chordwas 26.7 millions. The.transitionpoint on the lo_._ersurfacewas not dete._nined for ..... _e._..acJ_ythe foregoing condi%ions but,

....oe._i L_.tent o_:0.247 and as,s sho_._nin tab!o Iii_ at a lift _" _ '_"....I{eynolds number of 26.8 m_.].lion.str_nsltion occurred at 28.4.

percent of the Chord so %hat for C_, = 0.220, representing amore _,nqfavorablecondition for the #io_.zersurface, the extent ofthe i_mina_y_" layer _,;,oul.dbe somewhat ...._es,_"than =8.4o_percent ofthe chord, This res_!.t is ,,_indication of the de-gree of

,.m_.la_econdition as compared to thatinferiority of the io_,,_er _- "" _'of the best upper surface condition.

The influence of' surface condition on the position of

transition is shown .more directly by comparison between thetransition results obtained with the d,ifferont upper surfaceconditions. Wi,.,h condition A, for which th_ _nd_ated ampli--

•tude of the tr_sverso surface _av.lness was as much as 0,005

, inch, and with the ].eft engine ....,-.,_opped, transition occurredat 32.5 percent of the chord and 24 iDoho's outboard of thepanel center line at an airolano lift coefficient of 0.247 enda Reynolds number of'26.4 millions• For s_2"face condition D,with an indicatod waviness _mplitud.e of 0.001 inch, and thesame Dower_ condition the transition occurred at _,__0.4p_rcentof the chord at the seune ]._eynoldsnumber and a more unfavorablelift coe.[fi_ment of 0.256 The result of the improvement in theupp3r surface condition was th.orefore an increase in the extent

8

of the laminar boundary layer of at least lO percent of thechord. The effects of the intermedimte surface conditions ere

not definitely indicated by the re'_u!ts.

Operation of the engines and propellers h_d an advers_effect an the extent of the lamin_._ layer. Comparison of theres_D:bs obtained with both engines operatlng at f_fll throttlewith those obtained with both engines stopped indic,.wbess,reduction In the lamluar--flow r_m of about 3 percent of thechord,,

In flgu:ces 7 and 8 .... _ .... _ ....%..__n_a.rs-_.u,yez velocity distributions,determined .-'..or several co:<ditJons from the tests, are comparedwith the theore"bicai Biasius. flap-plate &istrib_zbions. In

general, the experimental points cc.-,.?ormto the theoreticalprofile shape within bhe prdbab}.c limits of scot_racy of themeasurements. The effect of'the favorable pressure ,_radient,which is __Inta.!ned over the fo_:_ard.45 percent of the

35,215 airf0ii section, i8 evidenced in.fl.g_re 7 by the values

of equivalent fl_b.-plate len..,_.h,corresponding to the Blaslusprofiles, which are generally less than the actual di-stancealong the stufface from the stagnation point.

The values of I_.5 derived :fro_:._lthe measured velocitydistributions in the l_..inar bous.dary layer end listed in

table IV range from about 7500 to 9000. A!th.ou':,31individualvalues may not be entirely reli_-zb]e_,the results, in ge_,ral,are sufficiently consistet_t to i_rmit the conclusion that

values of R_ of at least 8000 ar_ attainable before trs_,-sition occurs in flight on sui%£oly designed and caro.fullyfinished airfoils. The value 8000 represents a cons!der.:',ble

increase over the hi_hesb values obtained in the orlgir_alNACA lo_,-t_rbulenc,e tunnel on lamlnar-flow airfol].s similsr

to the j.,-_ .... sec_.fo_..,, this cbmp_rison indicates that ovenwith exbremely low turbulenco in the t_nmel air sbream_

botundary-layer and profiie-<Irag measurements may be subjectto considerable revision when applied to fiiL_t conditions.

It is po'inted out that while the value R= = 8000 may notbe the ultimate att_.inable, this value hasX been atbni-aed s]._dtherefore may be used as a guide in estimating what may beexpected in the o_fcent of gho lam.inar bounda:ry l__-s_yers_--d

hence in profile drag fo:_."airfoils havibg pressu_e-<listribu.tioncharacteristics .generally similar to ".:.hoseof the 35--215 airfoil.

The profile-drag coefficient of the panel was deterge.nedfrom the f.ttll-wakesur,zeys in accordance with the momentozamethod as developed by Jones. (See reference 3-) For the

9

power-off condition the coofficlent is m_hstantlally constantover the rankle of lift coefficient and _e_m..oidsnumber investi-gatod t_._l has a va.lue of about 0,00_.8, With po_;_T__on the va].u.eis increased to _,bouo 0_0052 or 8 _crcent.

In view of the inferior condition of the lower _urface of

the panel the profile-_d.ragm_as_.m'ementson bhe u,._er surfacealone are consi._leredas more nearly represontag.iv_ of the capa-bilities of the a._fo.i.i. The drag ooefficionts were eva_ ua_efrom the hs,lf-.,w_}<e _urveys by the mebhod of Saulre _(See reference ,._.)An shox,n in f'ig_re i0, for th/_ po:,_er-_oi_condition the coefficient Increm..Is@d f?oom 8]_out 0.002_ 8,t _._l

a_rp_ane l_It coefficient ef 0.2_ _d. a Reynolds n_mber of29,000_000 to 0.0028 at a lift co_ffic_'ent of 0.32 and aEe_nolds n_mber of 24,000,000. _,t is reasone,ble to a,_sr_e thatfor equally good soa_face conditions bhe drag due to the lowers_mfacs wot_ld be less than _hat of the uppe_ s_n'face so thatthe minimum drag coefficient of the airfoil would be somewhatless than 0.0044. The adverse effect on the drag coefficient

•duo to engine and propeller operation is substa__tiated by thepower-on results which show an increase in drag coefficientof abo_t lO percenb over tha _ower-o._f values

In reference 4_ in addition to the method, of determiningprofile dr_:_gfrom w_:e surveys, there is de,eloped a methodof predicting the drag from a knowledge of the !oc_tion ofthe transition point, the laminar botr_dary-layer velocityd_tribution i_medis,tel_ .,.c_,rd of the transition point,and the presst_re distribution between the transition pointand the trailing edge. To make use of this method the ex-perimenta.! pressure-41stributio_ curve for the upper surfacegiven in figure 5 was e_te_nded from 53 percent of the chordto the trailing edge where the pressure was known% from the half-wake su_sys. The profile-<Irag coefficient of the upper su._-x_._cewas then calc_f[ated for the cases of transition at 42 5percent and 32.5 percent of the chord, both at a Re2noldsnumber of _-_8,000,000. For the 42.5 p_rcent location the dragcoefficient was 0.0023 which is in close agreement with thevalue obtained by the wake-survey/ method. With trs:nsitionat 32.5 percent of the chord the drag coefficient, was calculatedto be 0.0028. Thesa results indicate a reduction of about

18 percent in the profile drag due to the improvement in su_-f_ce condition betwoen condition A and condition D.

The s]._..f_cance of the values of profile drag obtainedfrom the tests of the 35-215 airfoil section may become mor_apparent from suitable comparisons. For example, the theoret-ical turb_lent skln-friction dra{i_coefficient for two sides

i0

of a flat plate at the ReynOlds number n_t_rhich the •value of0.0048 was obtained for the test panel is 0.0052 or abo_t 8percent greater. The minimum profile-drag coefficient for theconventional NACA 0015 airfoil section is estimated to be0.00_57 at the same Reynolds number or about 20 percent greaterthan that of the 35-215 section. Comparison on the basis of

the uppeT' surface drag indicahes that _he single surfaceturbt_ent skin friction of a flat plate is about 12 percentgreater and the single sv_face drag of the 0015 section about30 percent greater than the upper s_rface drag of the 35-215airfoil section.

CONCLUDING IKE_r_RKS

A l_uuinar boundary layer was maintained over the uppersurface of the NACA 35-215 test panel to x/c = 0.424 wheretransition to turbulent flow occtu_red at a lift coefficientof 0.220 and a Reynolds number of 26_700,000. Improving thecondition of the upper surface so that the indicated amplitudeof the transverse _aves, as measured with the surface curvaturegage, was reduced from 0.009 inch to 0.00! inch resu/_ted inincreasing the extenb of the laminar bolmdary layer from 32.5perc,3nt to 42.5 percent of the chord_ thereby probably reducingthe profile-drag coefficient of the upper surface about 18 per-cont. The resltlts of the transition tests indicated a fo_ard

movement of the transition point of about 3 percent of the chorddue to operation of the engines and propellers.

(

The velocity slo:veys in the laminar bounda_j layer indicated

that values of bolnudary-layer Reynolds number R8 (based onthe distance from the st_face at which the dynemic pressure inthe boundary layerl is one-haJf that just outside the boundarylayer) exceeding 8000 are attainable in flight on suitablydesigned and carefully finished airfoils.

The profile-drag coefficient with power off was very nearlyconstant with a value of 0.0048 for flight conditions ranging

from s_%airplane lift coefficient of 0.21 and a correspondingReynolds ntmlber of about 30,000,000 to a lift coefficient of0.3 ° and a Reynolds number of 24,000,000. For the same r_geof conditions the profile-drag coefficient of the upper surfacealone varied from 0.0022 to 0.0028. The effect of full-throttle

operation of the engines and propellers increased the profile-drag coefficients as measured for both surfaces and for the

upper surface alone on the order of 8 to i0_percent.

Comparison of the res1_Its of the present flight tests onthe 35-215 airfoil, section with data obtained, on generallysimiJ.ar airfoils in the original NACA low-turbulence windt_mnol shod.Tedthat 5n fli.?_htthe l_in.ar bo_Idary layer was

maintained to values of I_$ considerably greater than thehi_hest values that were attained in the ttmnel. This resu].tindicated that oven in tu_mel air stresms of extremely lowturbulence the effect of the residual turbulence might be

appreciable, end thereby dmmonstra.ted the necessity of con-tinuod fli6ht resoa_?ch on airfoils of largo scale to supple ....merit the development work of the tunnels.

Langley Memorial Aeronautical L_ooratory_National Advisor,y Co-_:mmitteefor Aeronautics,

Ls_ag_l.eyField, Va., May 5._ 1941.

REFERENCES

i. Jacobs, Eastmsn N. : Proliminar._?Report on Laminar--FlowAirfoils and New Methods Adopted for Airfoil andBoundary-Layer Investigation. NACA ACR, Juno 1939.

2. The Cambrid_y_ University Aeronautics Laboratory: TheMeasurement of Profile Dra4_ by the Pitot-TraverseMethod. R <F_M No. ±sJoo,British A.R.C ].936.

3. Squire, H. B., and.Youa_g, A. D. : The Calculation of theProfile Drag of Iorofolls. R. £,_M. No. 1838_ BritishA.R.C.,. 1938.

TABZE2"O#D,'/vArEa.OF NACA 35-2/5 H/,eFo#_.

UPPE_ 5u_FAC-C Zo'wEeSuRFaCE

X IOD X/[O

0 0 0 0

/ ,oag 1.857 /.4/5 - 1.5_Y2.307 2._/,,o 21c)Y -e,/o.}4:. 78& 3.674- 5.2[4- -2.792.

7, 27_ 4.5/0 7. 7PE' -3322

_9,777 E.DII /0. D2J -3.752

14, 788 _,2,¢4- 15,DI2 -4,#4-B

1 I;9. 809 7, DPI _O,/91 -4.9,,_'7"_

24 8?8 7 6>._9 25 Id2 - =.. _.'-_ 9

22 673 8,_'g 90./27 -5.d_OQ,Z4.91-,¢ 8.774 9_,087 -5._8

. 9..9_,8 8. 9_1 40. Od? -.._9B,9

50. 07 ?" S>.702 49. _23 - 5-,762-

_0.150 Z 295 59,8ff0 -4. 709

70.IS'7 5.P77 u._.,,969 -_ 29_

DO.OSg 3.1D8 79.2/4 -/<817

r_r, O,_f 2 096' 6"4 944 - I. I_}0

90.022 1.175" ,, 89. 97/ - W61

95._o9 I , #?a ,94.9_J ! -_4,9I OO.oao _ 0 IO.O,oao 0

II

NATIONAL ADVI SORYCOMMITTE'E FOR AERONAUTICS

!

I NATION_&L ADVlSORT -

• 5u_ry of Re..c_/ts of Tx-_;_51_/o_ 7-es_'z o>7 Lower

Su/'.f_ce o_ AXA..C./_. _,-_-- Zl,.5- Tes_- Pa_vel ,Li,-_.

\ oF R _ c /rs 2:4-- /_ //_ e ,,_ c./_e._

i_!."zxl°° ',=-,,[R,___e of" A':'/,._,4-/-Conj." _r o_-,v _'o _ CL %_ R/fO_, CL %_ CL _] 0 @i A/o cL CL iR/ZO" cL

24.7 5 ,z65" 1 zZo .gSq _ Z#.3 I I ,Z_S'Iz,_.Z I

ZT.S 15-1,_o'7 I zg'_ ,al'71 z4,3 I z.,,-_ _.o I IZ8.4 7 ,ZII 1??,'7 ,'_15- I :Z.O,(=, I I I 2_9/ I Z4.S"30,3 4- .Z'¢_ IZ'7.3 ._t_l /'7,_ I I. .2'_/o'-lZ4,_ I

7 .ZII IZ'7,") ,4-l_lz_.(_ ,_ss i?-z.2 I .3o _/IZ3.8 i

5' ,z.38 Iz_.8 .Z_l z-_,_' 7" I 7- i

.z._-_ Iz?._ .ze8 1 z_.z- 7- I 7" I_. 2_./o ,z,'? I Z_ ,z_/ 12.4.8 7" I T I

I,¢ ,zB_- t 31._ z ,3Z_, I Z5-;4- 1 I .Z'_5" _Z_.5",_ I

13",Zo"? IZ_, 7 ,31"2 Iz4.3 I I ,3zzlz,;,-.z. i',_ ,Z_,P I Z_._ zy& I z_'.8 I I I 7"

38. / _' ,15'6 ] Zl,,,g ,z £,f' I ZG, z. I I I T

l'_o-'.z/'2 1_.9.5- .-ZG/ I Z_.8 I l I 7-

z,e._- 7 .z.$ �izT,e ,3:7,r" I zz,..y" I I I .z,_-71zt.a

'7 ,z3,/- iz?.B ,3"25-1 z_,.._ .z_j--Iz_.B I ,z_s- Iz5:.2 I8 ,z.iD I'Z_,o ..Z61 i z6,#- T I 7" I

3s.Z e z_,4 I_'_.,_ .3/_ I z4.8 T I T _I1o ,27_6 I z_Z I ,?-"2"2 I Zz,.5" 7" I T I

,Z'?.-_Iz_,,? ,3z3 Iz3,e I I ,Z2Z. IZ4.z- II¢

'?"_'_ Iz$"¢ ,Z_?I Z_S I .Z.?3 IZ_.? I

8 ,?__lO I_e,_ ,z_,i 126,4- I I I T38.1 e _z_,/- _ z_,/,, ,_/_, I z,_.8 I I I T

,/o ,2Z-& I z_?./ i,z?,_l z_,._ I I I T

_,a lq ,'z.za IZ_,.? i,_z3 IZ3. B I ,z_._lz3._" I I,z.5% IZG.s-i'zS'_ Iz_, 7" I I

Bot-_ E_9/_e5 S_opped

zB. �"7,z.se Ize.z Iz.z.z. I I I .z_olz_._,7 ,z3_ IZS.?_ ,_.o Izz.'z- Z_.5":l Z5:,8 I ,2:_,5"1Z5".8 I

8 ,Z¢6 I Z_.&" I,Z_'FI Z_.3 T I I i3_,z

1¢-,Zll Iz_,¢ ,Zo_Pa Iz_,2 I I .T I/....4- .Z/z. Iz.,_._' .Bzt IZ...*",5" I I ,3oo Iz_.8 I

38:/ B .Z,,_16 I Z;,5" ,Z"2 Z6.3 I I I 7"

Z5',_" "7 .Z_'¢, I Zr.o ,_33 l Z_.'z. I I I ,z73 IZ._..Y_,3.Z. "7 .Z_'_, !z{.o .4-J3 I Z_.'L ,3_'F!zz.3 I .z_lz_.j- . !

-/-l,e 9/,,_,_, k#_-,fe _-_ /'//34/'_#-_, C//-/'''_-'_J'.

L-532

NATIONAL ADVISORy

VALUE5 OF ici_6 DETERNIINED F-_OM BOUNDARY /TABLE .t7

/A Y_ MCASUREmENr5ON NAEA 35-215,'_I.eFo.;_ . SU_t4A_Y Cr PESU_r_ oF/_e_F-/LE D,_'AS

PowE_ SUrfAcE I -_xmo 3PANWJS_ r TC_T5oH /V_CA35-215A_R_L

UPPE,_SuRfACE P_wE_............ R//o

_OTH _-N61,_'E_ 32.5" CE_/TE_ .FTg BlO/-)

FURL- THrOTtLE C ._I B300 .20P Yl.5 . OOb-O

37.4- 12/NCHt$ . _11 e60(J .2.Y2 29. ? .D6_7OUTBOAI_D ,F./_ E'IO0 •256 27,3 ._U4_

• _0 _-_0 " _OT_ .256 283 .005_

32,._ CENTER .2_0 9700 _"NGI.VK5 25B 2/9.0 .0G49

.22L 7_0_ ETO_ED ,260 27,_ ,OP4@

Lp/ r ElVGl_/_- ,2_ e,DCO .2_O 27,5 .00_7

f T_'PP_D./ .25"_ ' 7?_'0

/Z--MLLT_OTTLF l_/,vc/_£j .2eb ' elOO .322 2_.3 _.... ,004_

_r_,4 OUTE_OA__ ._2_ _00 .214- 2.9."? .0_51

.2/s- e_'_o _OTH .249 27._ ._04)

_t. C CEtJTt_' . eTb 81b© /CULL " T_IR'._TT/.E ,282 26, I .0032

. _ ?_ •elO0 ' _/I 2"_,3 .0053

_/'_('TH _"-NE/NI _ C •e/J 846O L"P;ER ._U,_,:_CE ,4Z<JHE

JT.b'_I-_O " 12 I/VZ'/_., ¢ ,Z/_- 2aoo_7.4 .P22 _$00 ._Z'_ 27._ .b':.'2Y

Our _ox_O .22 7 22. 2 .po 20

• 77,_ 8COO

r D 40"P CE_VT"£_ ?IY i B_-&O_ _C'A/5/A/E5 .ebb 27.3 .OCk_

LOWEI_5"u_F,_CZ" .e?o D;.b ,ope_

• ._0.9 _ ?Oh' .2_3 24.5 . oo_

[ _6TH E/VGINL5 -- 3Y,2 CD,'¢Tt,_ .3/7 65g¢) ._2a _3,C_ .0;:_

i •]4a [,4PO BOTH _/=NG/IvE5 .226 - P? _ .00;°3

LEFT Ew6 j -- _3.2 CEI4TE,_ i.293 _oo __ __ " _

N A 'I*'JON AI# &D V r_OP)y

i